RTC Magazine

RTC:Motorola has long been identified as the leader in VME. Recently, there has been talk about some products going to end-of life, and Motorola’s commitment to VME has been questioned in some quarters. Could you set the record straight about VME vis-à-vis Motorola’s plans and product mix?

Vittori: The recent end-of-life announcement of our 68K-based products was necessary because of the discontinued availability of components needed to produce them. Perhaps a more significant point here was not the end-of-life announcement, but the fact that Motorola had continued to supply these 68K products for such a long period of time. This end-of-life announcement does not signify any change in our commitment to VME. We are offering product migration strategies that enable our customers to continue to use Motorola VME products for many years to come. Motorola has been the leader in VME since its inception in 1981. As we approach the 25-year anniversary of VME, we remain as committed as ever to maintaining our long-standing product leadership in VME.

RTC:Can we expect to see new VME product releases in the future, e.g., VITA 41, 46, etc.?

Vittori: VME is a key element of Motorola's Application-Enabling Platform strategy and we remain committed to leading the embedded computing industry with VMEbus products and technology. Our MVME6100 and recently announced MVME3100 demonstrate this product commitment and support for continued innovation such as 2ESST. We have also been very active in leading the definition of VXS standards. At this year's Bus and Board Show in January, we announced our support for VITA 41 and PCI Express fabric standards and it is our intent to develop product based on these standards. We do not currently have any plans to announce products based on the VITA 46 standard.

RTC: To some extent, ATCA is a relatively pure play in the communications sector, and VME largely fits into the industrial and military space. How do you see the mix of products—VME, CompactPCI and AdvancedTCA (ATCA) fitting into the various application areas and segments of the market? Do you anticipate any tie-in between the various form-factors—i.e., will Motorola be looking to launch ATCA, AMC and other form-factors into the industrial, commercial and military spaces?

Vittori: The AdvancedTCA standards were defined and developed by the PICMG 3.x standards committee, which took input from a broad spectrum of PICMG members. One of the critical objectives was to facilitate standards-based development in the telecommunications area, where the use of CompactPCI/PICMG 2.16 had already proven useful for applications at the network edge. Key trends we now see in the market include the convergence of traditional server-based applications and telecom applications. Global telecommunications competition is forcing service providers to carefully examine their business models with the objective of achieving rapid deployment of new services, while constraining costs to maximize revenue.

Layering standards-based software, such as high-availability middleware based on the SA Forum standards, onto standards-based hardware platforms, such as AdvancedTCA, gives equipment manufacturers many of the time-to-market and cost competitive advantages they need to meet carrier demands. This creates a horizontal industry model very much like that of the modern enterprise computing market, rather than the traditional, vertical, in-house focused development of the past. As a result, some segments of the enterprise computing market with reliability requirements that approach telecom levels are increasingly interested in solutions based on AdvancedTCA and SA Forum standards.

Smaller and more fragmented markets, such as the defense, medical and industrial automation segments, are still rooted in a vertical market model as demonstrated by the continuing strength of the market for VME boards. However, new programs such as the Navy Open Architecture and the Army/Air Force SOSCOE initiatives are more aligned with Motorola’s Application-Enabling Platform concept. An Application-Enabling Platform integrates industry-standard hardware and software elements into a computing platform that enables equipment manufacturers to deploy new products faster and with fewer resources. Motorola plans to offer Application-Enabling Platforms, based on VMEbus hardware, for this type of program.

There is increasing interest in the use of smaller, "low-profile" Advanced-TCA and MicroTCA platforms as an alternative to the large AdvancedTCA systems being deployed today. Smaller scale Application-Enabling Platforms can accelerate time-to-market, increase cost efficiency and enable customers to focus on their specific value-add in exactly the same way as our larger platforms. The interest in these technologies is coming not just from telecom equipment manufacturers, but also from the defense and industrial segments—especially where the application has a need for a high-availability solution.

RTC:  In the area of boards and systems, Motorola’s business model has been to market boards, but with the evolution of the industry and technology and particularly since the acquisition of Force, are there some cultural/business model evolutions going on as well?  Do you see Motorola moving toward offering more closely integrated computing platforms and if so, what are the implications for the industry in general? And, might Motorola contemplate taking that model into the VME space, providing system-level products?

Vittori: We started to change our business model about two years ago in anticipation of changes in the market. Many of our customers—the companies that design and build elements of the “Embedded Infrastructure”—share the same challenges:

• Competitive pressures that result in the need to continually introduce innovative, new products faster and at less cost. As these standards-based technologies increasingly perform core communications applications, the need for these building blocks to meet telecom five nines or higher reliability and serviceability requirements is increasing. Our old business model of providing a set of boards and other basic building blocks, which were then integrated by our customers, does not ensure that this need is met, or, there were significant time-to-market delays while our customers added these capabilities. That is why we are now focusing on providing a fully integrated, pre-validated application-enabling platform that is open to third-party content and is flexible, but also delivers to the telecom standards for reliability and serviceability on “Day 1.” The most significant benefits of this change will be to improve our customers’ time-to-market for new features and to significantly reduce both the development and product costs associated with more rapid technology insertion.

• The increasing complexity and importance of application software that provides differentiating product features—and the need to focus development resources on this added-value development. Some applications that were previously stand-alone now need to be seamlessly incorporated into a wider network. The combination of application software complexity, shorter time to market and development resource constraints means that preserving investment in application software is increasingly important. This includes re-use of application software across multiple products or programs running on different hardware platforms.

• Performance and environmental constraints such as increased processing power, real-time response and cooling limitations.

• Product lifecycle issues that can result in a continuous and costly development effort just to support existing product—effectively running fast to stand still. These include:

– The long service life of much embedded infrastructure

– The need for periodic technology upgrades to enhance performance.

– Electronic hardware components having shorter and shorter lives

– Continuous cost pressures on product support and upgrades

Our customers are addressing these issues by revamping their sourcing models, buying more and building less. They are relying on fewer trusted suppliers to provide more complete computing platforms and moving to common platforms that support a range of applications. Their traditional focus on hardware standards is being extended to a standards-based “common platform architecture.” This typically includes a base hardware, operating system and middleware on which applications can be built. The specific standards used vary based on industry and application needs.

Our highly integrated Application-Enabling Platforms are designed to help our customers be more competitive in their markets, increase cost efficiencies and reduce the risks inherent in developing new products. These benefits are very important for all of our customers—not just telecom equipment manufacturers—to help them respond to growing pressures in their own markets.

Our first Application-Enabling Platforms were designed for telecom applications. Motorola plans to extend the concept with platforms for other application areas—for example, the Navy Open Architecture and Army/Air Force SOSCOE initiatives. These will be based on VMEbus as well as other emerging form-factors such as MicroTCA.

RTC: In the matter of ATCA, several manufacturers seem to be ambivalent about how closely their ATCA products will be interchangeable. Will Motorola’s ATCA boards work in other chassis, or are we moving—for reasons either commercial or technical—away from the time that developers could mix and match standard products?

Vittori: Motorola firmly believes that the broad adoption of standards is an essential element for the future success of the telecom industry. All of Motorola's AdvancedTCA products are designed to be compliant with the appropriate PICMG 3.x standards and, as such, they will work in any AdvancedTCA-compliant chassis. Our strategy is to deliver Application-Enabling Platforms that integrate standards-based hardware and software to the industry. To be successful with such a model, Motorola must participate in a robust ecosystem of partners that can supply products that work in our platforms—this can only be effective with strict adherence to standards. 

When one takes a platform approach to a standards-based business, a number of different factors come into play. The AdvancedTCA standards include a number of different interconnection mechanisms which are: PICMG 3.1 Ethernet, 3.2 InfiniBand, 3.3 StarFabric, 3.4 Advanced switching and 3.5 Rapid IO. Although all blades from these different standards can be plugged into an AdvancedTCA shelf, only a set of blades with a single switch fabric standard can operate within a shelf. This may seem obvious, but the issue of multiple fabrics is important if AdvancedTCA is to reach its full economic potential.

Motorola defines the broad ecosystem of blades that are compatible with a switch fabric as a "blade set", and it is this blade set that really defines the scope of an effective ecosystem at the Application-Enabling Platform level. Virtually all systems deployed today are based on PICMG 3.1—Gigabit Ethernet. These are ideal for compute-centric, “low” bandwidth applications, such as those in the wireless market. To reach its full potential, AdvancedTCA needs to expand across the broad telecommunications market into applications where speeds of several hundred Gigabits per shelf are required. Adopting a faster fabric standard that is not backward compatible with those already deployed would create a new blade set ecosystem based around the faster standard. The more fabric technologies that are broadly adopted, the more fragmented the AdvancedTCA market becomes, preventing AdvancedTCA from achieving the economies of scale that the telecom industry needs.

Longevity and backward compatibility are essential attributes of any successful standard. The increasing deployment of Gigabit Ethernet (PICMG 3.1)-based systems means that there will be substantial industry pressure to deploy faster Ethernet speeds up to the point where Ethernet can deliver 10G (OC192, STM64) uplinks with traffic engineering/quality of service. Motorola is actively working with major Ethernet switch chip vendors to realize this potential in the marketplace over the next several years, all within the PICMG standards. This will enable the smooth upgrade of existing deployed systems and allow existing technology to evolve to support new high-speed systems, delivering the economies of scale necessary for industry success. 

RTC: There seems to be a number of maneuvers in the patent arena, specifically claims by Lucent over MicroTCA and AMC and now Motorola has filed applications on VITA 41 (VXS) and VITA 42 (XMC). Can you comment on the possible effects of this on the industry in general and, specifically, Motorola’s intentions?

Vittori: Motorola believes that the development and protection of intellectual property is a vital part of the standards/specification development process in protecting the business interests of all who participate. We further believe that the filing and defense of patents is the best method of doing so.

We affirm our commitment to abide by the intellectual property policies of the standards/specification development organizations (SDOs) in which we hold memberships, among them the VITA Standards Organization, PICMG, Service Availability Forum and Open Source Development Laboratory.

Motorola's Embedded Communications Computing business has a long history of creating intellectual property to enhance the overall value of the industry, and we will continue to offer fair, reasonable and non-discriminatory licensing terms for any Motorola IP that promotes the development of a vigorous open ecosystem of suppliers.

In the specific case of VXS and XMC, we will commit not to assert our rights to recently disclosed IP unless assertions are made against us. We are doing this in advance of the issuance of the patents we have applied for because we believe that it sets a healthy precedent for our colleagues in the industry.

RTC:   In the mainstream press there are numerous articles on how Comcast, AOL/Time Warner, Cox and other cable companies will be offering VoIP telephone service to more than 40 million customers in coming months. As the world moves closer to VoIP, what platforms do you see leading, and how do you see the demands on platforms evolving?

Vittori: Everyone realizes that simply replacing the Plain Old Telephone Service with the same thing based on IP is not a good business decision—that business model probably won't be sustainable in the long term. Carriers are looking to attract consumers by offering more and better services in a faster and more cost-effective manner, and this is where Motorola's vision of "seamless mobility" can help. Once basic voice traffic is in the form of IP, then it can be processed and converted easily to match a wide range of target personal devices. By using our Application-Enabling Platforms, telecom equipment manufacturers that supply the carriers can enable a faster build-out of the infrastructure that underpins seamless mobility and many of these new services.

CPCI-based systems are already being installed alongside, or to replace, conventional exchange switches and deliver VoIP and other broadband services. Going forward, this application set is likely to migrate to AdvancedTCA, which is seen to offer better upward scalability. However, remote sites and enterprises require smaller scale switching platforms with lower entry cost and here the emerging MicroTCA is likely to play a key role. Enterprise-focused blade servers, such as IBM eServer BladeCenter, will be strong in VoIP Server applications, such as Media Servers and Voice Response systems used in enterprises or ISP environments.

RTC:  There are those who predict that ATCA may go the way of contract manufacturing, i.e., commoditization. Intel, for example, is already offering volume-produced processor blades. Do you believe that ATCA can avoid becoming a commodity as volumes pick up? Is it reasonable to think of ATCA as yet another standard form-factor supplied by the merchant market board vendors, or do we need to develop another way of looking at it and the systems built around it?

Vittori: To be successful AdvancedTCA needs to reach high volume deployment. This can only be accomplished by minimizing the number of standards used—minimizing the number of "blade sets" and competing successfully in appropriate markets with blade server and related technology, from which it already faces challenges. Otherwise, AdvancedTCA, as a hardware standard, faces relegation to niche applications in telecom, defense and related markets. The emergence of these more cost-effective communications platforms will allow for much more widespread deployment of communications computing capabilities than has been possible in the past, due to the high cost of these products. Successful participants in the AdvancedTCA market segment, both suppliers and users, will benefit from these more attractive cost/volume economies. We also anticipate a virtuous “cycle of innovation” in which new products, with ever-increasing price-performance advantages, will be introduced on a continuous basis. So, it is not just a matter of introducing one set of products, which then become commoditized. There will, in fact, be more and more opportunities for innovative suppliers to introduce new products that are not commodities. What is so exciting about this change is that we will see a much more vibrant technology innovation environment than we have seen before in the traditional embedded computing market segment.

There is another important change to embedded computing that is happening at the same time. The adoption of standards-based software running on top of standards-based hardware is a much more powerful concept than just a hardware standard. The Service Availability Forum has defined a set of Application Programming Interfaces (APIs) that provide standards-based mechanisms for accessing basic hardware information (fan speeds, blade temperatures, card insertion/removal etc.) and a set of APIs for high-availability middleware services, such as event handling, locking, message distribution, clustering (in the sense of applications/hardware that are related to each other) etc. The integration of standards-based middleware and a standards-based hardware platform into an Application-Enabling Platform creates the dynamic of a horizontal business model in markets such as telecom and this approach will enable AdvancedTCA to thrive in a highly competitive environment.

RTC:  What is Motorola’s software strategy in terms of Linux, major RTOSs, Java and standards like the Service Availability Forum (SAF) interfaces?

Vittori: Motorola believes that the Service Availability Forum standards are critical to future success in many major markets. SAF's Application Interface Specification (AIS) creates a standard for high-availability middleware that has broad market application, in both the telecom and enterprise markets. Motorola's SA Forum-compliant middle-ware implementation includes the concept of portability across both hardware platform architectures and blade-level operating systems. This will allow big endian and little endian processor architectures to reside in the same platform using a variety of operating systems.

Motorola believes that in markets where AdvancedTCA will thrive and compete, Linux will be the predominant operating system. It is ideal for soft real-time applications, such as those found in the telecom world. Motorola is working with MontaVista, the leading supplier of carrier-grade Linux in this area. However, our strategy is not ultimately limited to a single Linux vendor or a single operating system. Applications that require hard real-time operating systems can also operate in Motorola's SA Forum-compliant middleware environment.

Motorola believes Java will play an important role in telecom applications in the future. Java is increasingly used in the enterprise world as an easily implemented environment that enables simple applications to operate on available data. The SA Forum, which Motorola currently chairs and to which Motorola heavily contributes, has just begun efforts to incorporate a Java model into the SA Forum Application Interface Specification layer. This will eventually enable a broad array of Independent Software Vendors (ISVs) to develop high-availability-capable Java applications that can easily be deployed across a wide variety of platforms—telecom, enterprise and related industry segments, such as defense and medical. This creates a very efficient horizontal model very similar to the server market today.

RTC:  Could you please let us know what you believe about the future and market for CompactTCA? MicroTCA? AMC? XMC?

Vittori: AMC technology is definitely seen as the next extension of mezzanine technology and offers hot swap and non-disruptive upgrade capabilities that are very interesting to telecom as well as many other industries. The larger form-factor and power envelope compared with existing PrPMC technology is also attractive. However, mezzanine technology incurs a cost simply due to the additional connectors and traces that are required, which will result in cost/performance tradeoffs versus full-sized AdvancedTCA blades and competing server technology. MicroTCA seeks to leverage AMC technology into a backplane interconnect to deliver an efficient smaller form-factor implementation. This can be attractive in highly cost-sensitive markets or space constrained markets. Node B/Base Transceiver Stations and ETSI front access 300 mm depth shelves are examples of areas in the Telecom market where MicroTCA is attracting significant interest, even though standards are yet to be developed.

The key point is that all these platforms can support a SA Forum-compliant high-availability middleware layer to create an off-the-shelf Application-Enabling Platform, enabling users to migrate applications from one technology to another more easily than ever.